Global Journal of Health Science; Vol. 8, No. 5; 2016 ISSN 1916-9736 E-ISSN 1916-9744 Published by Canadian Center of Science and Education

Psychometric Testing of the Greek Version of the Clinical Learning Environment-Teacher (CLES+T) Evridiki Papastavrou1, Maria Dimitriadou1 & Haritini Tsangari2 1

School of Health Sciences, Department of Nursing, Cyprus University of Technology, Cyprus

2

School of Business, Department of Economics and Finance, University of Nicosia, Cyprus

Correspondence: Maria Dimitriadou, School of Health Sciences, Department of Nursing, Cyprus University of Technology, Cyprus. Tel: 357-9967-1898; Fax: 357-25-002-864. E-mail: [email protected] Received: June 1, 2015 doi:10.5539/gjhs.v8n5p59

Accepted: June 24, 2015

Online Published: August 31, 2015

URL: http://dx.doi.org/10.5539/gjhs.v8n5p59

Abstract Clinical practice is an important part of nursing education, and robust instruments are required to evaluate the effectiveness of the hospital setting as a learning environment. The study aim is the psychometric test of the Clinical Learning Environment+Teacher (CLES+T) scale-Greek version. 463 students practicing in acute care hospitals participated in the study. The reliability of the instrument was estimated with Cronbach’s alpha coefficients. The construct validity was evaluated using exploratory factor analysis (EFA) with Varimax rotation. Convergent validity was examined by measuring the bivariate correlations between the scale/subscales. Content, validity and semantic equivalence were examined through reviews by a panel of experts. The total scale showed high internal consistency (α=0.95). EFA was identical to the original scale, had eigen values larger than one and explained a total of 67.4% of the variance. The factor with the highest eigen value and the largest percentage of variance explained was “supervisory relationship”, with an original eigenvalue of 13.1 (6.8 after Varimax rotation) and an explanation of around 38% of the variance (or 20% after rotation). Convergent validity was examined by measuring the bivariate correlations between the scale and a question that measured the general satisfaction. The Greek version of the CLES+T is a valid and reliable instrument that can be used to examine students’ perceptions of the clinical learning environment. Keywords: clinical learning environment, nurse education, psychometric testing, validation 1. Introduction Nurse education is an integral part of any healthcare delivery system as it has a major role to play in the educational development of graduates who can deliver safe, good quality nursing care. Nursing professionals’ knowledge, attitudes and skills are acquired both through formal education in institutions and through experience in the clinical area (Gaberson & Oermann, 2010). The clinical area represents an environment where the student nurse interact emotionally, physically and cognitively with their surroundings and fulfill their learning outcomes. Under these circumstances the clinical area is referred to as the ‘clinical learning environment’. Clinical learning environment (CLE) is viewed as an invaluable resource in familiarizing students with the reality of their professional role (Henderson, Cooke, Creedy, & Walker, 2012). Students being exposed to a range of clinical experiences (Edwards, Smith, Courtney, Finlayson, & Chapman, 2004) helps to broaden and deepen both cognitive and psychomotor skills (Souza, Venkatesaperumal, & Radhakrishnan, 2013), develop caring relationships and aid nurses in their professional evolution, as embodied within the attitudes of the nursing workforce (Haugan, Sørensen, & Hanssen, 2012). Interest in the concept of clinical education as a determinant of quality nursing has gained increasing attention since 1980, leading to various nursing education reforms (Pollard, Ellis, Stringer, & Cockayne, 2007). Historically, a number of researchers investigated the effectiveness of the clinical learning environment (CLE) from students’ perspectives, employed quantitative (Saarikoski, Marrow, Abreu, Riklikiene, & Özbicakçi, 2007; Melender, Jonsén, & Hilli, 2013; Papathanasiou, Tsaras, & Sarafis, 2014; Dimitriadou, Papastavrou, Efstathiou & Theodorou, 2015), qualitative (Papp, Markkanen, & Von Bonsdorff, 2003; Mattila, Pitkäjärvi, & Eriksson, 2010) and mix methodology (Bisholt, Ohlsson, Engström, Johansson, & Gustafsson, 2014; Ip & Chan, 2005). The pedagogical atmosphere characterized by respect, acceptance and opportunities for learning with the mentor and clinical teacher alike have a stake in making clinical learning successful and reliable (Papp et al., 2003; 59

www.ccsenet.org/gjhs

Global Journal of Health Science

Vol. 8, No. 5; 2016

Saarikoski et al., 2007; Dimitriadou et al., 2015). Also, student satisfaction was consistently identified as the most reliable index of a “good” clinical learning environment (Chan, 2001; Papastavrou, Lambrinou, Tsangari, Saarikoski, & Leino-Kilpi, 2010; Dimitriadou et al., 2015). For that reason, a variety of instruments for evaluating CLE from students’ perspective have been developed such as the CLEI-Clinical Learning Environment Inventory (Chan, 2001); the CLESDI-Clinical Learning Environment Diagnostic Inventory scale (Hosoda, 2006); the BES-CPE (Levett-Jones, 2009); the SECEE - Student Evaluation of the Clinical Education Environment (Sand-Jecklin, 2009); and the CLES - Clinical Learning Environment and Supervision Scale. The CLES is a 27-item scale (Saarikoski & Leino-Kilpi, 2002) developed to measure both the clinical learning environment and clinical supervision. Later, this scale was reviewed by the developers and a new sub-dimension added culminating in a 34-item CLES + T scale (Saarikoski, Isoaho, Warne, & Leino-Kilpi, 2008) and assesses 5 factors. They are as follows- pedagogical atmosphere, leadership style of the ward manager, supervisory relationship, the premises in the ward, and the role of the nurse teacher. The CLES+T instrument was used mainly and extensively in Europe in an effort to develop a powerful, multilingual tool for evaluating the quality of clinical learning. To date, the CLES+T scale has been translated into nine different languages, i.e., English, Finnish, Italian, Greek, Swedish, Dutch, Norwegian, German and Spanish. It has been demonstrated to be a valid and reliable tool among different international samples (Saarikoski et al., 2008; Johansson et al., 2010; De Witte, Labeau, & De Keyser, 2011; Henriksen, Normann, & Skaalvik, 2012; Tomietto et al., 2012; Bergjan & Hertel 2013; Saarikoski et al., 2013; Watson et al., 2014; Vizcaya-Moreno, Pérez-Cañaveras, De Juan, & Saarikoski, 2015). Since the quality of clinical learning depends on how well the practice curriculum is structured, the use of a reliable tool is helpful in obtaining information regarding the effective organization of clinical practice and the quality of it’s supervision. With this in mind, CLES+T is selected to test the psychometric properties with the aim to enhance cross-cultural collaboration. 2. Methods 2.1 Aim The current study aim was to test the psychometric properties of the Greek version of the Clinical Learning Environment Scale + Teacher (CLES+T-GR) of Greek Cypriot students. 2.2 Participants and Settings The target population of the study included all nursing students enrolled at the three Universities offering a Bachelor’s degree in Nursing within Cyprus. Only students who were practicing in hospitals were recruited, and those who were practicing in primary health care centers and other community care settings were excluded. 664 students fulfilled the inclusion criteria and were given the questionnaire. 463 questionnaires were returned, giving a response rate of 70.33%. 2.3 Ethical Considerations and Data Collection The research proposal was submitted to the National Bioethics Committee, which is responsible for all research projects in Cyprus according to the Law (3558/2001). The permission for access to the field research was obtained by the Chairs of Nursing from each university. The aims of and the rationale for the study, and assurances that the data would be processed anonymously were included in the information letter supplied with the questionnaire. The questionnaires were given to the students individually immediately after they had completed their clinical placement, during a nursing laboratory lesson. After completion, each questionnaire was returned in a closed envelope. The completion of the questionnaire was considered as an informed consent for participation in the study. 2.4 Research Instrument As mentioned the CLES+T scale consists of 34 items classified into 5 subscales: (1) pedagogical atmosphere on the ward; (2) supervisory relationship; (3) leadership style of the ward manager; (4) premises of ward nursing; and (5) role of the nurse teacher in clinical practice (Saarikoski et al., 2008). Respondents are asked to score their perception to each item on a 5-point Likert-type scale ranging from “very dissatisfied” to “very satisfied”. Also demographic data, hospital and ward type, length of clinical placement, number of meeting with the nurse teacher, motivational level on clinical placement, and level of satisfaction were collected. 2.5 Data Analysis For demographic data and scale items, descriptive statistics (frequencies, percentages, means and standard deviations), skewness and kurtosis were used. The internal consistency of the Greek version of the instrument and each dimension was estimated with Cronbach’s alpha coefficients. Also, item analysis was conducted on the data, providing item-to-total correlations and Cronbach’s alpha if the item was deleted from the scale. Construct 60

www.ccsenet.org/gjhs

Global Journal of Health Science

Vol. 8, No. 5; 2016

validity was evaluated using exploratory factor analysis (EFA) with Varimax rotation. Principal components analysis (PCA) was implemented as the extraction method in EFA. Convergent validity was examined by measuring the bivariate correlations between the scale/subscales and the question about the general satisfaction of nurses. Content, validity and semantic equivalence were examined through review by a panel of experts of the content of each item, its wording and the meaning of the items, after translation, in the context of the Cypriot culture (Squires et al., 2013). 3. Results 3.1 Sample Characteristics The final sample included 463 participants. Among those, 38.7% were males and 61.3% females, with ages ranging from 18 to 34 years, a mean of 21.08 years and standard deviation 2.23 years. 149 participants studied in private universities, and 314 in the single public university. 3.2 The Individual Scale Items Descriptive statistics (mean and standard deviation) for individual items were calculated. In order to examine the variability of the answers and test for significant deviations from normality, item skewness and kurtosis were also reported. Acceptable values of skewness and kurtosis, based on George and Mallery (2001), are those between -1.5 and +1.5, whereas values between -1 and +1 are considered excellent. The highest mean was of an item on the supervisory relationship subscale (item 18 in the scale - My mentor showed a positive attitude…), with a mean of 4.3, and a (low) standard deviation of 1.00. This item had marginal values of skewness and kurtosis as well. This verified that most of the answers were “agree” or “fully agree”, as opposed to negative attitudes, but the values were within the acceptable range. Although, in most of the items, there was a weak trend towards the positive attitudes (agree and strongly agree), there were no critical values (high positive or negative) of kurtosis or skewness for any item with all values being in the acceptable range. Therefore, transformations were not deemed necessary and all items were included in the analysis. These results are presented in Table 1. 3.3 Internal Consistency and Reliability-Item Analysis Cronbach’s alpha was used for testing the reliability of the 34-item scale, as well as for the five subscales, where values close to one are considered satisfactory. Also, item analysis was conducted to provide information about how well each individual item correlated to other items in the sub-scale where corrected item-to-total correlations below 0.30 are usually considered unacceptably low (Polit & Beck, 2011). Also, all of the inter-item correlations were examined with the reliability of each item being considered by finding if the it’s Cronbach’s alpha was deleted. The results showed high internal consistency for the total scale (α=0.95). Similarly, the reliability of each sub-category was found to be high, ranging from 0.81 (“nursing care”) to 0.96 (“supervisory relationship”). The latter results are presented in table 1. Item analysis showed that when any item was deleted from the scale, the alpha was slightly lower or approximately the same as compared to when all the items were included, suggesting that deleting any item does not change the overall reliability significantly, and as such, that all the items contribute to the high reliability of the scale. A slight increase was seen in items “I felt comfortable going to the ward at the start of my shift” and “the staff learned how to know the students by their names”, however small (from 0.949 to 0.950). The items can thus be considered reliable. Corrected item-to-scale correlations varied from 0.38 to 0.71, showing that all correlations were satisfactory, that is, above 0.3. Finally, out of more than 250 inter-item correlations between the 34 items (not reported), only two correlations exceeded 0.80 (in the supervisory relationship factor). Thus, in general, the results suggest that within the scale no items duplicated each other. The corrected item-to-scale correlations and Cronbach’s alpha if the item was deleted appear in Table 1.

61

www.ccsenet.org/gjhs

Global Journal of Health Science

Vol. 8, No. 5; 2016

Table 1. Descriptives, Internal consistency and Reliability (sequence as presented in the questionnaire) Item

Mean

Std. Deviation

Skewness

Kurtosis

Corrected Item-Total Correlation

Cronbach Alpha if item Deleted

Pedagogical Atmosphere, 9 items, Cronbach’s alpha 0.875 The staff were easy to approach

3.7646

1.06635

-.843

.183

.522

.948

I felt comfortable going to the ward at the start of my shift

3.0606

1.22013

-.159

-.893

.385

.950

During staff meetings(e.g. before shifts) I felt comfortable taking part in the discussions

4.0475

1.03296

-1.160

1.093

.493

.948

There was a positive atmosphere on the ward

3.8013

1.04210

-.840

.426

.620

.948

The staffs were generally interested in student supervision

3.2505

1.19427

-.248

-.733

.605

.948

The staff learned to know the students by their personal names

2.4514

1.31973

.500

-.870

.403

.950

There were sufficient meaningful learning situations on the ward

3.4190

1.08981

-.356

-.525

.620

.948

The learning situations were multi-dimensional in terms of content

3.3826

1.08331

-.289

-.449

.613

.948

The ward can be regarded as a good learning environment

3.6609

1.09773

-.508

-.476

.613

.948

Leadership style in ward management, 4 items, Cronbach’s alpha 0.849 The WM regarded the staff on his/her ward as a key resource person

3.9391

1.03774

-.970

.620

.546

.948

The WM was a team member

3.5502

1.18464

-.628

-.351

.490

.949

Feedback from the WM could easy be consider a learning situation

3.3275

1.13437

-.323

-.525

.584

.948

The effort on individual employee was appreciated

3.3348

1.09000

-.286

-.425

.531

.948

Nursing care, 4 items, Cronbach’s alpha 0.812 The ward nursing philosophy was clearly defined

3.6811

.98259

-.569

.090

.576

.948

Patients received individual nursing care

3.8304

1.02769

-.696

-.099

.512

.948

There were no problem in the information flow related to patients’ care

3.7609

1.01158

-.636

-.015

.471

.949

Documentation of nursing ( e.g. nursing plans, daily recording of nursing procedures etc.) was clear

3.8824

.99743

-.730

.129

.589

.948

Supervisory Relationship, 8 items, Cronbach’s alpha 0.849 My supervisor showed a positive attitude towards supervision

4.2673

1.05593

-1.482

1.504

.658

.947

I felt that I received individual supervision

3.7733

1.18752

-.734

-.326

.647

.947

62

www.ccsenet.org/gjhs

Global Journal of Health Science

Vol. 8, No. 5; 2016

I continuously received feedback from supervisor

3.9399

1.16111

-.857

-.219

.683

.947

Overall I am satisfied with the supervision I received

3.9690

1.22435

-1.057

.110

.705

.947

The supervision was based on a relationship of equality and promoted my learning

3.9498

1.15256

-.921

-.048

.703

.947

There was a mutual interaction in the supervisory relationship

3.9569

1.14764

-.987

.251

.714

.947

Mutual respect and approval prevailed in the supervisory relationship

4.0597

1.11053

-1.098

.456

.681

.947

The supervisory relationship was characterized by a sense of trust

4.0072

1.14320

-1.001

.173

.703

.947

Role of nurse teacher, 9 items, Cronbach’s alpha 0.937 In my opinion, the NT was capable of integrating theoretical knowledge and everyday practice of nursing

3.8366

1.18799

-.834

-.179

.607

.948

The NT was capable of operational sing the learning goals of this placement

3.8293

1.18870

-.855

-.107

.601

.948

The NT helped me to reduce the theory-practice gap

3.6800

1.22331

-.704

-.389

.543

.948

The NT was like a member of the nursing team

3.4181

1.27918

-.417

-.804

.501

.949

The NT was able to give his or her expertise to the clinical team

3.5730

1.19551

-.561

-.455

.550

.948

The NT and the clinical team worked in supporting my learning

3.4900

1.22629

-.442

-.718

.573

.948

The common meetings between myself mentor and NT were comfortable experience

3.8491

1.15296

-.794

-.174

.594

.948

In our common meetings I felt that we are colleagues

3.8604

1.14160

-.776

-.232

.632

.947

Focus on meetings was in my learning needs

3.5643

1.25568

-.553

-.699

.552

.948

3.4 Content Validity and Semantic Equivalence The Greek Version of the CLES-GR was translated and back-translated following a specific step procedure (Papastavrou et al., 2010) after obtaining consent from the authors. Although the content validity of the Greek version of the CLES-GR has been established, the questions were reviewed by five experts as the questionnaire was modified by the designers and re-named to include the nurse teacher. The expert panel agreed that the CLES+T-GR reflected the situation in the clinical practice environment, i.e., that the items were suitable and relevant to be tested on Cypriot students and it was of acceptable face validity. 3.5 Construct Validity Construct validity was evaluated using exploratory factor analysis (EFA) with Varimax rotation. Principal components analysis (PCA) was implemented as the extraction method in EFA. The procedure used is similar to the psychometric studies published for the CLES+T (Saarikoski et al., 2008; Johansson et al., 2010; Henriksen et al., 2012; Bergjan & Hertel, 2013; Vizcaya-Moreno et al., 2015). First, the assumptions regarding the suitability 63

www.ccsenet.org/gjhs

Global Journal of Health Science

Vol. 8, No. 5; 2016

of the data for factor analysis were examined, including the sample size, Kaiser–Mayer–Olkin (KMO) measure of sampling adequacy and Bartlett’s test of sphericity. The sample size was satisfactory, considering the rule-of-thumb for determining a priori sample size to be a “participant- to- item” ratio of 10:1 (Costello & Osborne, 2005). The current study, included 34 items and 463 participants, therefore this ratio was satisfied. The data were found to be appropriate for factor analysis, since the KMO measure was equal to 0.931, larger than 0.5, indicating high sampling adequacy, whereas Bartlett’s test of sphericity was significant (p

Psychometric Testing of the Greek Version of the Clinical Learning Environment-Teacher (CLES+T).

Clinical practice is an important part of nursing education, and robust instruments are required to evaluate the effectiveness of the hospital setting...
NAN Sizes 0 Downloads 6 Views